1. Field of the Invention
The present invention relates to a network access control method, a network system using the method, and apparatuses configuring the system, and, in particular, to a network access control method for an IP network, a network system using the method, and apparatuses configuring the system for an IP network
2. Description of the Related Art
As a network access control method in the related art, there is an access control method employing a proxy server. The proxy server is software or a server machine for repeating accesses to various services of Internet such as multimedia database, WWW (World Wide Web) and so forth.
The proxy server is used as a cache function in a case where traffic is to be reduced between a network in a company and the Internet. For example, when a user in the company accesses a page of WWW, the contents obtained therefrom is stored by the proxy server for a predetermined period. Then, when another user accesses the same page, it is not necessary to access the information through the Internet, and the information stored by the proxy server is transferred to this other user. Thereby, it is possible to reduce the frequency of access operations to the Internet.
Further, there is another access control method of previously setting the number of connections by a server application. For example, Japanese Laid-Open Patent Application No. 5-316115 discloses such a method. In this method, a plurality of computer networks each connecting a plurality of computers are mutually connected, and, through access control made by using information of date, day of the week, time, and/or group number, load distribution for a private line will be performed precisely and efficiently. In the method, for this purpose, a previously set information table (including date, day of the week, time, group number, and so forth) is referred to, and thereby, access control for the networks is performed.
Further, as disclosed in Japanese Laid-Open Patent Application No. 6-152615, there is a network access control method for the purpose of performing efficient uniform load control. In the method, a function of measuring a load of a transmission path and a function of switching a transmission path are provided, and, thereby, a load (traffic) of each transmission path is measured and a transmission path having a low load is selected.
Further, a load balancing system has been also proposed. In this system, for the purpose of distributing a load of a server to which traffic is concentrated, a plurality of mirror servers are provided, load situations of the servers and mirror servers are measured, and the mirror series are caused to access the server which has a lowest load.
In the above-mentioned method employing the proxy server, in a case where an object of an information transmitting source is updated after a content object is once stored in a cache, the object is not updated because the substance of the object is not accessed for the subsequent access operations. Accordingly, no new information can be obtained every time an access request is made. Further, a dynamic object (using Common Gateway Interface, Server Side Include, Active Server Pages, and so forth) which varies in contents each time access is made as a result of search by a search engine cannot be re-used. Further, the cache function has a limitation of a memory capacity, disk capacity, available term of cached information, or the like. Accordingly, in some case, even an object which is regarded as effective should be erased when a situation in the capacity is serious and confusing.
In the above-mentioned method of previously setting the number of connections, the user should access many times when a traffic is concentrated. Accordingly, a useless traffic flows through a network. Further, depending on the previously set information (date, day of the week, time), or even in a case where a traffic is measured, and a path having a low load is selected, path selection may not result in a sufficient effect unless a server to which a connection is made has a sufficient processing capability.
In the above-mentioned method of measuring of load situations of servers and mirror servers, and causing the server having a low load to be accessed, because this method merely distributes accesses from users, access control is not performed in a case of a heavy load applied to each server or each mirror server. For the server to which traffic is concentrated, merely the number of mirror servers is increased uselessly unless access regulation is made by the end of the network.